Time-dependent Reliability And Sensitivity Analysis For Mechanical Structure

Reliability is closely related to the mechanical structure. The traditional reliability analysis methods often deal with the problem without considering the strength degradation under varying loads and accurate model. Thus, the reliability result is a constant value. However, the actual engineering problems are often very complex. At the same time, because of the actual structure's changing from reliable state to failed state gradually, an intermediate transition fuzzy area exist between the reliable state and failed state. The reliability analysis method should give full consideration to these issues. The main work is as followed:1. The time-dependent structural reliability analysis method under the fuzzy failure criterion can be used to solve the reliability problems with the load and resistance varying over time. In this method, the performance function under the fuzzy failure criteria is discretized, then apply the basic theory of fuzzy sets and introduce new random variables to get the equivalent performance function in different time periods. Eventually the time-dependent structural reliability problem is simplified into the static probabilistic reliability problem.2. Based on the TRPD model, a sensitivity analysis method is proposed to deal with the time-dependent reliability problems in the complex engineering structures. The method can be used to obtain the sensitivity indices, which are used for the sensitivity of the failure probability with respect to the random sensitive variable parameters. Equations of these sensitivity indices were derived. Three examples were used to demonstrate the accuracy and effectiveness of the proposed method.3. The improved PHI2 method for structural time-dependent reliability analysis is based on the computation of the outcrossing rate. Firstly, transforming the time-dependent reliability problem into a conventional time-dependent system reliability problem. Secondly, linearizing the limit-state function with the first-order reliability method. Thirdly, applying the basic theory of the PHI2 method and furthermore, introducing a new random variable and two parameters, the converted system reliability problem can be efficiently solved.